Presentation is loading. Please wait.

Presentation is loading. Please wait.

Group Analyses Guillaume Flandin SPM Course Zurich, February 2014

Similar presentations


Presentation on theme: "Group Analyses Guillaume Flandin SPM Course Zurich, February 2014"— Presentation transcript:

1 Group Analyses Guillaume Flandin SPM Course Zurich, February 2014
Wellcome Trust Centre for Neuroimaging University College London With many thanks to W. Penny, S. Kiebel, T. Nichols, R. Henson, J.-B. Poline, F. Kherif SPM Course Zurich, February 2014

2 Statistical Inference
Image time-series Spatial filter Design matrix Statistical Parametric Map Realignment Smoothing General Linear Model Statistical Inference RFT Normalisation p <0.05 Anatomical reference Parameter estimates

3 GLM: repeat over subjects
fMRI data Design Matrix Contrast Images SPM{t} Subject 1 Subject 2 Subject N

4 First level analyses (p<0.05 FWE):
Data from R. Henson

5 Modelling all subjects at once
Fixed effects analysis (FFX) Modelling all subjects at once Simple model Lots of degrees of freedom Large amount of data Assumes common variance over subjects at each voxel Subject 1 Subject 2 Subject 3 Subject N

6 Modelling all subjects at once
Fixed effects analysis (FFX) Modelling all subjects at once Simple model Lots of degrees of freedom Large amount of data Assumes common variance over subjects at each voxel = +

7 Within-subject Variance
Fixed effects Only one source of random variation (over sessions): measurement error True response magnitude is fixed. Within-subject Variance

8 Random effects Two sources of random variation: measurement errors
response magnitude (over subjects) Response magnitude is random each subject/session has random magnitude Within-subject Variance Between-subject Variance

9 Random effects Two sources of random variation: measurement errors
response magnitude (over subjects) Response magnitude is random each subject/session has random magnitude but population mean magnitude is fixed. Within-subject Variance Between-subject Variance

10 Probability model underlying random effects analysis
𝜎 𝑤 2 𝜎 𝑏 2 Probability model underlying random effects analysis

11 Fixed vs random effects
With Fixed Effects Analysis (FFX) we compare the group effect to the within-subject variability. It is not an inference about the population from which the subjects were drawn. With Random Effects Analysis (RFX) we compare the group effect to the between-subject variability. It is an inference about the population from which the subjects were drawn. If you had a new subject from that population, you could be confident they would also show the effect.

12 Fixed vs random effects
Fixed isn’t “wrong”, just usually isn’t of interest. Summary: Fixed effects inference: “I can see this effect in this cohort” Random effects inference: “If I were to sample a new cohort from the same population I would get the same result”

13 Terminology Hierarchical linear models: Random effects models
Mixed effects models Nested models Variance components models … all the same … all alluding to multiple sources of variation (in contrast to fixed effects)

14 = + Hierarchical models = + Example: Two level model Second level
First level

15 Hierarchical models Restricted Maximum Likelihood (ReML)
Parametric Empirical Bayes Expectation-Maximisation Algorithm But: Many two level models are just too big to compute. And even if, it takes a long time! Any approximation? spm_mfx.m Mixed-effects and fMRI studies. Friston et al., NeuroImage, 2005.

16 One-sample t-test @ second level
Summary Statistics RFX Approach fMRI data Design Matrix Subject 1 Subject N First level Second level One-sample second level Contrast Images Two-stage procedure Generalisability, Random Effects & Population Inference. Holmes & Friston, NeuroImage,1998.

17 Summary Statistics RFX Approach
Assumptions The summary statistics approach is exact if for each session/subject: Within-subjects variances the same First level design the same (e.g. number of trials) Other cases: summary statistics approach is robust against typical violations. Might use a hierarchical model in epilepsy research where number of seizures is not under experimental control and is highly variable over subjects. Balanced designs Mixed-effects and fMRI studies. Friston et al., NeuroImage, 2005. Statistical Parametric Mapping: The Analysis of Functional Brain Images. Elsevier, 2007. Simple group fMRI modeling and inference. Mumford & Nichols. NeuroImage, 2009.

18 Summary Statistics RFX Approach
Robustness Summary statistics Hierarchical Model Listening to words Viewing faces Mixed-effects and fMRI studies. Friston et al., NeuroImage, 2005.

19 ANOVA & non-sphericity
One effect per subject: Summary statistics approach One-sample t-test at the second level More than one effect per subject or multiple groups: Non-sphericity modelling Covariance components and ReML

20 GLM assumes Gaussian “spherical” (i.i.d.) errors
sphericity = iid: error covariance is scalar multiple of identity matrix: Cov(e) = 2I Examples for non-sphericity: non-identically distributed non-independent

21 2nd level: Non-sphericity
Error covariance matrix Errors are independent but not identical (e.g. different groups (patients, controls)) Errors are not independent and not identical (e.g. repeated measures for each subject (multiple basis functions, multiple conditions, etc.)) Clinical populations same subjects on placebo, drug1, drug2

22 2nd level: Variance components
Error covariance matrix Cov(𝜀) = 𝑘 𝜆𝑘𝑄𝑘 Qk’s: Qk’s: Clinical populations same subjects on placebo, drug1, drug2

23 Example 1: between-subjects ANOVA
Stimuli: Auditory presentation (SOA = 4 sec) 250 scans per subject, block design 2 conditions Words, e.g. “book” Words spoken backwards, e.g. “koob” Subjects: 12 controls 11 blind people Data from Noppeney et al.

24 Example 1: Covariance components
Two-sample t-test: Errors are independent but not identical. 2 covariance components Qk’s: Error covariance matrix

25 Example 1: Group differences
controls blinds First Level Second Level design matrix

26 Example 2: within-subjects ANOVA
Stimuli: Auditory presentation (SOA = 4 sec) 250 scans per subject, block design Words: Subjects: 12 controls Question: What regions are generally affected by the semantic content of the words? “turn” “pink” “click” “jump” Action Visual Sound Motion Noppeney et al., Brain, 2003.

27 Example 2: Covariance components
Error covariance matrix Errors are not independent and not identical Qk’s:

28 = = = Example 2: Repeated measures ANOVA ? ? ? First Level Second
Motion Sound Visual Action First Level ? = ? = ? = Second Level

29 ANCOVA model Mean centering continuous covariates for a group fMRI analysis, by J. Mumford:

30 Analysis mask: logical AND
16 12 8 4

31 SPM interface: factorial design specification
Options: One-sample t-test Two-sample t-test Paired t-test Multiple regression One-way ANOVA One-way ANOVA – within subject Full factorial Flexible factorial

32 Summary Group Inference usually proceeds with RFX analysis, not FFX. Group effects are compared to between rather than within subject variability. Hierarchical models provide a gold-standard for RFX analysis but are computationally intensive. Summary statistics approach is a robust method for RFX group analysis. Can also use ‘ANOVA’ or ‘ANOVA within subject’ at second level for inference about multiple experimental conditions or multiple groups.

33 Bibliography: Statistical Parametric Mapping: The Analysis of Functional Brain Images. Elsevier, 2007. Generalisability, Random Effects & Population Inference. Holmes & Friston, NeuroImage,1998. Classical and Bayesian inference in neuroimaging: theory. Friston et al., NeuroImage, 2002. Classical and Bayesian inference in neuroimaging: variance component estimation in fMRI. Mixed-effects and fMRI studies. Friston et al., NeuroImage, 2005. Simple group fMRI modeling and inference. Mumford & Nichols, NeuroImage, 2009.


Download ppt "Group Analyses Guillaume Flandin SPM Course Zurich, February 2014"

Similar presentations


Ads by Google